8
$\begingroup$

I'm wondering what types of devices exist that can detect how close someone is to falling asleep. I have heard of several devices that may be able to do this, but I haven't been able to find very solid information.

  • EEG: I assume these could detect falling asleep quite effectively; though, I'd like to know if there are any other devices that would work as well.
  • EKG/ECG: I found this paper that claims you can use a Lorenz plot to detect correlation between heartrate and falling asleep.
  • GSR: A lot of people say a GSR can be used, but I have yet to find any papers/data that describe the correlation between GSR and falling asleep
  • Pulse Oximeter: someone mentioned to me that these work as well, though I haven't been able to find much information
$\endgroup$
5
  • $\begingroup$ pupillometry and/or PERCLOS $\endgroup$
    – Jeff
    Commented Mar 28, 2013 at 4:45
  • $\begingroup$ @Jeff That's quite interesting; though, I was thinking more along the lines of someone who's laying in bed, trying to go to sleep. In that case, they'd usually have closed eyes the whole time. $\endgroup$
    – Azmisov
    Commented Mar 28, 2013 at 5:22
  • $\begingroup$ Do we actually know what happens in the brain exactly at sleep onset? Is there actually a definite "border" from the perspective of brain activity? When I re-awaken shortly after having fallen asleep, I remember continuous thoughts, and only waking up makes the difference clear. What happens, though, is loss of conscious muscle control. Depending on sleeping position, you can recognize muscles loosening, and on the verge of sleep people often jerk when they notice their limbs moving after they release their "hold" and react to still this movement. So I would measure muscle tone and/or twitch $\endgroup$
    – user1196
    Commented Mar 28, 2013 at 6:09
  • $\begingroup$ Problem with many measures (pupils, muscles, brain activity) would be to discern sleep onset from other states with similar patterns (meditation, arousal [widening of pupils] etc.). $\endgroup$
    – user1196
    Commented Mar 28, 2013 at 6:12
  • 2
    $\begingroup$ @what Muscle tone only drops off completely in REM, when there is a gating between circuits in the brainstem and spinal cord (see this Bio.SE question). The American Academy of Sleep Medicine has guidelines for sleep onset based on EEG frequencies on certain leads which were originally put forth in a seminal work in the 1960's (Rechtschaffen and Kales), see this for a comparison between the two systems. $\endgroup$ Commented Mar 29, 2013 at 4:47

2 Answers 2

5
$\begingroup$

The gold standard of sleep studies is polysomnography- using multiple sensors together to score sleep stages. It's very expensive and requires a lab. A night in a sleep lab costs about 3000$, last time I checked.

Actigraphy is a cheaper way of monitoring sleep by looking at the users motion. When 19 out of 20 minutes are scored as "no activity", the first minute of the 20 minutes sequence is classified as sleep onset. such Actigraphy algorithms that detect sleep onset latency have to 87% accuracy compared with polysomnography, and at very low cost. Chances are your iPhone or android device can be used as an Actigraphy to detect sleep onset. I wrote one such app for android and 2 for iPhone. Zeo is similar, although it works also by detecting motion related artifacts due to sleep stages, maybe as much as real brain waves.

$\endgroup$
7
$\begingroup$

I've used the Zeo (http://www.myzeo.com/sleep/). It seems to work pretty well. It tracks which stage of sleep you are using a very basic EEG. There are some details on how it operates here: http://radar.oreilly.com/2009/10/how-the-zeo-sleep-device-works.html

As I said, I've used it in the past and from what I can tell, it seems to be pretty accurate, assuming you don't knock the sensor off while you're sleeping.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.